Week 3: What is a temperature?

One molecule in an empty box has a speed and an energy. It does not have a temperature. Add an enormous number of them and temperature becomes one of the most reliable things you can measure.

10 August 2026

Put one molecule in a sealed, empty box, and ask what its temperature is.

You can ask plenty of reasonable questions about that molecule. How fast is it moving, and in which direction? How much kinetic energy does it carry? Each of those has a definite answer at any instant. Temperature is not one of them. A single molecule does not have a temperature in the ordinary thermodynamic sense. It has a speed, a position, an energy — properties that belong to it alone, with no thermometer reading that makes sense for one particle bouncing around an otherwise empty box.

Now open a valve and let in a mole of gas: somewhere near ten to the twenty-three molecules, each one doing exactly the same kind of thing the lone molecule was doing — moving, colliding, changing direction, carrying momentum and energy.

Suddenly temperature is one of the most reliable numbers in the room. Put a thermometer in the gas and it settles on a value. Check it again a minute later, undisturbed, and it has not drifted. Two boxes of the same gas at the same temperature behave identically in every way that matters for thermodynamics, regardless of which specific molecules happen to be where at any moment.

Nothing new was added to the box except more of the same thing that was already there. No molecule turned into a different kind of object between the first case and the second. So where did this new, well-behaved property come from?

Not from any single molecule. It comes from the distribution — the spread of speeds and energies across the whole population. A gas at a given temperature has molecules moving at wildly different individual speeds, from nearly still to far above average, but the shape of that spread is stable and predictable. For a gas in thermal equilibrium, temperature is encoded in that distribution. It measures something about the collective statistics of an enormous number of degrees of freedom, not something carried by any one of them.

That is worth sitting with. Position and velocity are properties every molecule has on its own, whether or not anyone is looking. Temperature is not like that. It only becomes meaningful once there are enough particles for a distribution to exist at all, and even then it is a statement about the whole ensemble, never about one member of it picked out and interrogated alone.

Last week’s glow was called thermal because its spectrum had the shape a system at a definite temperature produces. That claim can now be read for what it actually says. At the moment light stopped scattering, the universe’s radiation had a distribution of energies with the same statistical shape a gas settles into in an ordinary box. No single particle carried that temperature. The whole population did.

Which raises a harder question than it looks like at first. If a property like temperature exists only for the crowd and for none of its members, in what sense is it real?

What we explained#

One molecule in a box has a speed and an energy but not a temperature. Add enough molecules and temperature becomes one of the most reliable quantities you can measure, even though nothing new was added except more particles of the same kind. Temperature belongs to the statistical distribution of the whole population, not to any single member of it.

What became stranger#

How does a crowd have properties its members don’t?